Using public databases for genomic prediction of tropical maize lines

We showed that it is possible to use public datasets as a primary training set and that population structure can modify the predictive abilities of genomic selection

Pedro Patric Pinho Morais


Scholarcy highlights

  • Proposed at the beginning of the twenty-first century by Meuwissen, Hayes, and Goddard, genomic selection arose as a broad methodology for using information from markers spread over the entire genome to predict the performance of genotypes
  • The addition of lines belonging to the São Paulo University panel panel in the optimized training set and randomized training sets groups led to an increase in the estimates of predictive ability for all the number of inbred lines in the training tested
  • TA B L E 5 Mean and standard deviation f predictive ability) of the genomic predictions obtained in training set group 4 for different sizes of the optimized training set obtained via the USP, North Central Regional Plant Introduction Station and Association panel panels for the traits of plant height and ear height
  • There is a decrease in predictive ability, Nt = 1,500 and 500, for PH and EH, respectively. These results show that the structure of the population plays a vital role in optimizing the training sets because when the population effect is smaller, the OTSs can more accurately predict the lines of the USP panel
  • The results obtained via training set group 3 and TSG4 indicate that predictive ability can be improved if the lines that compose the training group are correctly selected by efficient methods, such as that proposed by Akdemir et al, making satisfactory results feasible mainly through smaller groups of individuals
  • We recommend the use of optimization methods, as an example Akdemir et al, to build the training sets
  • Small groups of individuals selected from public panels are sufficient to achieve satisfactory, over 0.53, for predictive abilities of genomic selection

Need more features? Save interactive summary cards to your Scholarcy Library.